| Follow @yukisww
Los Angeles
Yuki is a Software Engineer in the Cloud Infrastructure team at Edmunds, where he builds highly resiliant and performant systems on the AWS platform. There, he leads the open source project ShadowReader, a serverless load testing framework being used every day at Edmunds. In his day to day, he works with technology like Docker, Kubernetes, Spinnaker, and Serverless.
Authored Comments
Thank you, that's a very interesting comment. It wasn't outlined in this post, but at Edmunds there is a Master lambda invoked for each application being tested. So, there could be an App1 Master and App2 Master running in parallel, each running their own Worker lambdas with different sets of URLs. When load testing the entire Edmunds website, over 30 Masters are invoked concurrently by the Orchestrator. Hope that makes it more clear why we went with this approach.